Research Paper

Neurochemical correlates of rapid treatment response to electroconvulsive therapy in patients with major depression Stephanie Njau, BA; Shantanu H. Joshi, PhD; Randall Espinoza, MD; Amber M. Leaver, PhD; Megha Vasavada, PhD; Antonio Marquina, PhD; Roger P. Woods, MD; Katherine L. Narr, PhD

Background: Electroconvulsive therapy (ECT) is a highly effective brain stimulation treatment for severe depression. Identifying neurochemical changes linked with ECT may point to biomarkers and predictors of successful treatment response. Methods: We used proton magnetic resonance spectroscopy (1H-MRS) to measure longitudinal changes in glutamate/glutamine (Glx), creatine (Cre), choline (Cho) and N-acetylaspartate (NAA) in the dorsal (dACC) and subgenual anterior cingulate cortex (sgACC) and bilateral hippocampus in ­patients receiving ECT scanned at baseline, after the second ECT session and after the ECT treatment series. Patients were compared with demographically similar controls at baseline. Controls were assessed twice to establish normative values and variance. Results: We included 50 patients (mean age 43.78 ± 14 yr) and 33 controls (mean age 39.33 ± 12 yr) in our study. Patients underwent a mean of 9 ± 4.1 sessions of ECT. At baseline, patients showed reduced Glx in the sgACC, reduced NAA in the left hippocampus and increased Glx in the left hippocampus relative to controls. ECT was associated with significant increases in Cre in the dACC and sgACC and decreases in NAA in the dACC and right hippocampus. Lower NAA levels in the dACC at baseline predicted reductions in depressive symptoms. Both ECT and symptom improvement were associated with decreased Glx in the left hippocampus and increased Glx in the sgACC. Limitations: Attrition and clinical heterogeneity may have masked more subtle findings. Conclusion: ECT elicits robust effects on brain chemistry, impacting Cre, NAA and Glx, which suggests restorative and neurotrophic processes. Differential effects of Glx in the sgACC and hippocampus, which approach control values with treatment, may reflect previously implicated underactive cortical and overactive subcortical limbic circuitry in patients with major depression. NAA levels at baseline are predictive of therapeutic outcome and could inform future treatment strategies.

Introduction Major depressive disorder (MDD) is a debilitating illness that affects approximately 350 million people worldwide.1 There is accumulating evidence to support that the pathophysiol­ ogy of MDD involves multiple overlapping biological sys­ tems, the contributions of which may vary across individ­ uals.2 In pursuit of more effective treatments at the individual level, there is thus a critical need to objectively identify bio­ markers of disease and treatment response.3 Electroconvul­ sive therapy (ECT) is a rapidly acting and highly effective treatment for relieving severe depression in patients for whom standard pharmacological interventions have failed.4 Despite impressive remission rates for ECT, with estimates ranging from 75% to 90%,4,5 there are potential cognitive side effects. Further, similar to other antidepressants, there is a

lack of knowledge and consensus regarding the mechanism of antidepressant action.6 Cumulative evidence describes depression as a brain net­ work disorder likely driven by abnormalities in corticolimbic circuits, including hippocampal and amygdalar regions and the dorsal (dACC) and subgenual (sgACC) anterior cingulate cortex and connected prefrontal cortex (PFC).7–9 For example, structural alterations in the PFC, particularly the ACC, and in the hippocampus are consistently observed in patients with major depression.10–12 Similarly, alterations in white matter pathways connecting these regions have been reported.7,8,13 Recent studies suggest these same circuits are impacted by treatment.14–16 Further, work from our group has shown changes in hippocampal volume with ECT 17 as well as changes in dACC neurochemistry18 and in white matter path­ ways linking prefrontal–limbic regions. 19 These findings

Correspondence to: K. Narr, 635 Charles E Young Dr S #225, University of California Los Angeles, Los Angeles CA 90095; [email protected] Submitted May. 18, 2015; Revised July 30, 2015; Accepted Jan. 31, 2016; Early-released June 23, 2016 DOI: 10.1503/jpn.150177 © 2017 Joule Inc. or its licensors

6

J Psychiatry Neurosci 2017;42(1)

Treatment and brain chemistry in major depression

s­ uggest the dACC, sgACC and bilateral hippocampus play key roles in the pathophysiology of MDD. Various neuroimaging techniques have been increasingly applied to better understand depression and treatment-­ related neuroplasticity in implicated frontolimbic regions. Proton magnetic resonance spectroscopy (1H-MRS) is a noninvasive imaging technique that can identify neurochemical signals in brain tissue, such as glutamine/glutamate (Glx), N-acetylaspartate (NAA), phosphocreatine-creatine (Cre), choline and choline-containing compounds (Cho).20 As such, 1 H-MRS offers a unique opportunity for probing neurochem­ ical correlates and predictors of treatment response. Though not an extensive literature, several prior 1H-MRS studies have addressed links between altered neurometabolite concentra­ tions and depression.20 Most consistently reported are obser­ vations of reduced glutamate levels in the ACC.18,21,22 One study reported a reduced Glx:Cre ratio within the hippocam­ pus in unmedicated patients with unipolar MDD.23 Studies exploring differences in NAA indicate reductions in the NAA:Cr ratio for the thalamus24 and reduced NAA for the caudate,25 yet no differences have been reported previously in the ACC, PFC, amygdala or hippocampus.20 At least 1 re­ port shows elevated Cho in the putamen.25 Early evidence suggests that ECT modulates altered brain chemistry in patients with MDD. For example, compatible with 2 earlier reports,21,22 our group has shown an increase, indicating normalization toward control values, of glutamate levels within the ACC in patients with MDD after 6 ECT ses­ sions.18 Though most investigations of hippocampal NAA changes with ECT have not found differences,26 at least 1 study has reported ECT-related increases.27 Another study reported a significant increase in hippocampal Cho levels ­after ECT in depressed individuals.26 Despite some prior research, owing to small sample sizes and a focus on mostly single and varied anatomic regions, it remains unclear whether neurochemical markers are associ­ ated with and/or predict antidepressant response in regions that form components of both cortical and subcortical limbic pathways, which play pivotal roles in mood regulation and emotion. For metabolites showing treatment effects, it is also unclear whether signal changes are a consequence of the brief seizures elicited by ECT or whether they are more closely tied to the neurobiology underlying improvements in clinical symptoms. Furthermore, the question of whether neuro­ chemical profiles are predictive of clinical response has been mostly unexplored. To address the questions above, we sought to examine 1H-MRS–derived changes in neurochemis­ try underlying ECT treatment and ECT-related clinical re­ sponse and to establish whether metabolite levels measured before ECT predict subsequent treatment outcome. We exam­ ined patients with MDD before, during and after ECT and compared them cross-sectionally with demographically simi­ lar nondepressed controls. Controls were assessed twice to estimate normative values and variance. Based on the exist­ ing literature, we hypothesized that ECT would be associated with a normalization of Glx (change in the direction of con­ trols) in the ACC and that disease-related changes in NAA, Cho and Cre would also vary with ECT response. To our



knowledge, this is the largest and most comprehensive study to date to address the links between 1H-MRS markers and rapid response to ECT in patients with MDD in both dorsal and ventral limbic pathways simultaneously.

Methods Participants We recruited patients with unipolar or bipolar depression who were experiencing a major depressive episode and who were scheduled to receive ECT at the University of Califor­ nia, Los Angeles (UCLA) Resnick Neuropsychiatric Hospital. Diagnosis was determined by a board-certified psychiatrist (R.E.) following DSM-IV-Revised criteria and confirmed ­using the Mini-International Neuropsychiatric Interview (M.I.N.I.). To be included, patients had to have experienced 2  or more earlier major depressive episodes and failed to ­respond to at least 2 prior medication trials. We excluded pa­ tients with separately diagnosed comorbid psychiatric disor­ ders, including schizophrenia, schizoaffective disorders, posttraumatic stress disorder, bipolar mania, attention-­ deficit/hyperactivity disorder and dissociative disorders, and those with primary anxiety disorders. Other exclusion criteria were dementia, first-episode depression, illness onset after age 50 years, depression related to serious medical ill­ ness, or any neuromodulation treatment (e.g., vagal nerve stimulation, repetitive transcranial magnetic stimulation) within 6 months of the ECT treatment index series. All pa­ tients were tapered off psychotropic medications, including antidepressants and benzodiazepines, over a period of 48– 72 hours before enrolment and ECT treatment. Patients were scanned and administered the Hamilton ­Rating Scale for Depression (HAMD-17) at 3 time points: baseline (T1; occurring within the 24  hours preceding their first ECT session), first follow-up (T2; occurring between the second and third ECT treatment sessions) and second followup (T3; occurring within 1 week of completing the ECT treat­ ment index, about 4 wk after the first treatment). We used advertisements to recruit controls matched for age and sex from the same geographical area as patients. Controls were administered the M.I.N.I. to exclude those with a history of depression, other psychiatric or medical illness and/or a history of antidepressant use. Controls were scanned at 2 time points: baseline (C1) and follow-up (C2; occurring approxi­ mately 4 wk after the initial scan). Exclusion criteria for all participants were history of alco­ hol or substance abuse within the 6 months preceding the study and/or substance dependence within the 12 months preceding the study, any neurologic disorder and contraindi­ cation to MRI scanning. All participants provided written in­ formed consent, and the UCLA Institutional Review Board approved our study protocol.

ECT treatment During the index series, ECT (5000Q MECTA Corp.) was ad­ ministered 3 times a week using a standard protocol for

J Psychiatry Neurosci 2017;42(1)

7

Njau et al.

a­ nesthesia (1  mg/kg of methohexital) and paralysis (1  mg/ kg of succinylcholine). Administration of ECT followed the seizure threshold (ST) titration method, where after establish­ ing the ST, treatments were delivered at 5 × ST for right uni­ lateral (RUL) d’Elia lead placement using an ultrabrief pulse width (0.3 ms) and at 1.5 × ST for bilateral placement using a brief pulse width (0.5 ms).

Clinical measures The HAMD-1728 and Montgomery-Åsberg Depression Rating Scale (MADRS)29 were administered at the same time points as brain scanning. Because these scales are highly correlated, HAMD-17 ratings were chosen as the primary measure of clinical response and used to determine associations with 1 H-MRS markers.

Imaging protocol Single-voxel point resolved spectroscopy (PRESS) sequences were acquired using a Siemens 3 T Allegra system (repetition time [TR] 2200  ms, echo time [TE] 30 ms, spectral width 2000  Hz, 1024 samples) with and without water suppression (128/1 averages). We used a volumetric navigator to correct for motion and B0 inhomogeneities in real time.30 High-­ resolution motion-corrected multi-echo MPRAGE images31 were acquired using the same scanner (TEs 1.74, 3.6, 5.46 and 7.32 ms; TR 2530 ms; inversion time [TI] 1260 ms; flip angle 7°; field of view 256 × 256 mm; 192 sagittal slices; voxel reso­ lution 1.3 × 1.0 × 1.0 mm3) and resliced to position 1H-MRS voxels of interest (30 × 12 × 12 mm) in the midsagittal dACC and sgACC and in the left and right hippocampal grey mat­ ter (20 × 18 × 12 mm; Fig. 1). Table 1 shows the mean fullwidth at half-­maximum (FWHM) and signal-to-noise ratios estimated for each voxel.

Data preprocessing To improve the signal-to-noise ratio, we first used a Gauss– Seidel iterative update scheme to denoise 1H-MRS signals from each voxel.32 Water-referenced metabolite concentra­ tions were then computed with LCModel software. For Glx, metabolite concentrations from data points with Cramer–Rao lower bounds equal to or exceeding 20% were excluded from analysis. For NAA, Cre and Cho, Cramer–Rao lower bounds equal to or exceeding 10% were excluded, as these metabo­ lites typically display lower fitting error values. For cerebro­ spinal fluid (CSF) correction, preprocessing included removal of nonbrain tissue from the T1 data and field inhomogeneity correction using BrainSuite software.33 Thereafter, voxels of interest were mapped to the T1 space in which they were in­ itially prescribed. We used the FSL stats tool to compute voxel tissue composition (grey mater, white matter and CSF in each voxel). Finally, to control for potential differences in tissue compositions between participants and experimental groups, metabolite concentrations were corrected for the frac­ tion of CSF content in each voxel.34

Statistical analysis For the metabolites of interest (Glx, NAA, Cho and Cre), sta­ tistical analyses addressed cross-sectional differences at base­ line between patients and controls, longitudinal effects of ECT, correlations with clinical response using HAMD-17 ­ratings, and baseline metabolite levels as predictors of treat­ ment outcome. For the analysis of cross-sectional differences at baseline, we used analyses of covariance to explore differences includ­ ing age and sex as covariates. For the analysis of longitudinal effects of ECT, we used the general linear mixed model (GLMM), as it produces unbiased parameter estimates when

Fig. 1: Example voxel placement and LCmodel processing output for the subgenual anterior cingulate cortex, dorsal anterior cingulate cortex and right hippocampus (voxel placement for the left hippocampus is similar to the right and is not shown).

8

J Psychiatry Neurosci 2017;42(1)

Treatment and brain chemistry in major depression

Results

observations are missing at random. It therefore allowed the inclusion of all data points in spite of missing cell values as a result of exclusionary Cramer–Rao bounds or missing time points. The GLMM used an unstructured covariance matrix, and time point (T1, T2 and T3) was included as a continuous variable with random intercepts and slopes to account for within-subject correlations. For metabolites showing signifi­ cant main effects of ECT, we then examined time points pair­ wise. The GLMM was also used to compare metabolite val­ ues across time points (C1 and C2) in controls to establish normative variance for repeated measures. Focusing only on the metabolites showing significant lon­ gitudinal effects of ECT, the analysis of correlations with clin­ ical response used the general linear model (GLM) to analyze the difference scores between baseline and the end of the ECT index in order to establish associations between change in metabolite levels and change in HAMD ratings. The analy­ sis of baseline metabolite levels as predictors of treatment outcome used the GLM to address whether baseline metabo­ lite levels predicted overall clinical response (i.e., the differ­ ence in HAMD ratings between T1 and T3). In contrast to the GLMMs, which included all available data points, only par­ ticipants completing each of the 3 time points for whom dif­ ference scores could be computed were included in our analy­ses of correlations with clinical response and baseline metabolite levels as predictors of treatment outcome. Follow-up analyses compared patients’ T3 measures to controls’ T1 measures to establish whether metabolites show­ ing ECT effects in patients normalize or diverge from control values over the course of treatment. Finally, we performed post hoc analyses to test for poten­ tial interactions between patients with unipolar or bipolar de­ pression as well as for the effects of lead placement (defined as the percentage of ECT sessions including RUL) for those measures showing significant ECT effects. Because lead placement was clinically determined and not manipulated in this study, effects of lead placement were examined while controlling for baseline HAMD ratings. Because some early evidence, including our own work, suggest that ECT elicits changes in 1H-MRS signals of Glx, NAA, Cho and Cre,18,20,22 we set a 2-tailed α level of 0.05 as the significance threshold to determine main effects of ECT. To reduce potential type 1 error, only metabolites showing significant main effects of ECT were subsequently examined for effects of joint longitudinal change with mood ratings and predictive effects.

Demographic and clinical variables Fifty patients (43 with unipolar, 7 with bipolar depression; 23  men, 27 women) and 33 controls (14 men, 19 women) matched for age and sex participated in the study. Figure 2 de­ scribes the flow of participants through the study and scanning sessions, and the demographic and clinical characteristics of participants are shown in Table 2. Patients and controls did not differ in sex (χ21,82 = 0.35, p = 0.55) or age (F1,82 = 2.19, p = 0.14). In patients, HAMD and MADRS ratings improved significantly with ECT (F2,12.20 = 30.05 and F2,37.24 = 30.04, both p < 0.001).

Cross-sectional effects of diagnosis Prior to ECT, patients displayed reduced left hippocampus NAA (F1,67 = 11.38, p = 0.001) and elevated left hippocampus Glx (F1,63 = 4.03, p = 0.049) compared with controls. In contrast, for the sgACC, Glx levels were decreased in patients relative to con­ trols at baseline (F1,53 = 5.35, p = 0.025; Fig. 3). Means for each ­metabolite by group and time point are provided in Table 3.

Longitudinal effects of ECT The GLMM analysis showed main effects of ECT in the dACC, where patients exhibited decreased NAA (F2,32.09 = 6.99, p = 0.003) and increased Cre (F2,29.30 = 7.67, p = 0.002) over the course of treatment. We further observed ECT-­ related reductions in left hippocampal Glx (F2,39.29 = 6.62, p = 0.003) and right hippocampal NAA (F2,49.33 = 7.31, p = 0.002). Significant effects of ECT were also present for Glx (F2,27.89 = 3.35, p = 0.05) and Cre (F2,26.65 = 5.48, p = 0.010) in the sgACC, with both showing increases over the course of ECT. For 1 H-MRS markers showing main effects, time points were compared pairwise (T1 v. T2, T2 v. T3, T1 v. T3). Probability values for pairwise comparisons are included in Figure 3. Metabolite values did not differ across time in control par­ ticipants (all p > 0.05; Table 3).

Associations between ECT-related clinical response and metabolite concentrations Correlations between changes in metabolite concentrations and HAMD scores indicated decreases in Glx in the left hip­ pocampus (F1,29 = 4.898, p = 0.035) and increases in Glx in the

Table 1: Mean full-width at half-maximum and signal-to-noise ratio for each voxel Mean ± SD Region

Signal-to-noise ratio

FWHM, ppm

No. of data points excluded with NAA SD > 10%

Dorsal ACC

18.03 ± 3.78

0.05 ± 0.01

9

Subgenual ACC

13.68 ± 3.68

0.09 ± 0.03

34

Left hippocampus

14.71 ± 3.54

0.09 ± 0.02

18

Right hippocampus

14.03 ± 3.64

0.09 ± 0.02

15

ACC = anterior cingulate cortex; FWHM = full width at half-maximum; NAA = N-acetylaspartate; SD = standard deviation.



J Psychiatry Neurosci 2017;42(1)

9

Njau et al.

sgACC (F1,19 = 4.15, p = 0.05; Fig. 4) occurred with improved mood ratings.

Predictors of treatment outcome Higher baseline metabolite levels of NAA in the dACC (F1,27 = 4.22, p = 0.029) predicted greater changes in HAMD scores between T1 and T3 (Fig. 5).

Follow-up analyses Follow-up analyses comparing T3 metabolite values in pa­ tients with T1 metabolite values in controls (to explore pres­ ence or absence of normalization, i.e., changes in the direc­ tion of controls) showed a significant difference for NAA (F1,57 = 10.79, p = 0.002) and Cre (F1,57 = 4.19, p = 0.045) in the dACC and for Cre in the sgACC (F1,45 = 8.40, p = 0.006).

Effects of electrode placement and diagnostic group No interactions were observed for lead placement or diagno­ sis of unipolar or bipolar depression (all p > 0.05) for metab­ olites showing significant longitudinal effects of ECT.

Discussion Several models with varying degrees of clinical and preclin­ ical evidence, most commonly the neurotransmitter, neuro­ endocrine, anticonvulsant and neurotrophic models, are pro­ posed to explain the rapid antidepressant effects of ECT.6,35 To clarify the underlying mechanisms, the current investiga­ tion sought to identify the effects of ECT on 1H-MRS markers of brain chemistry and the correlates and predictors of thera­ peutic response. For this purpose we used single-voxel 1 H-MRS, which is typically associated with a higher signalto-noise ratio than multivoxel sampling, to determine changes in metabolites in the dACC and sgACC and in the left and right hippocampus, which are regions repeatedly im­ plicated in MDD and its treatment.10–12,14–16 Overall, our find­ ings demonstrate that ECT elicits robust but varied effects on neurochemistry within these components of frontolimbic net­ works. Specifically, results showed decreased NAA and in­ creased Cre in the dACC, increased Glx and Cre in the sgACC, and decreased right hippocampal NAA and left hip­ pocampal Glx with ECT in patients, whereas all metabolite levels remained stable across time in controls. At baseline, patients differed from controls for Glx in the sgACC and left hippocampus and for NAA in the left hippocampus. Baseline NAA levels in the dACC were shown to predict ECT-related clinical response in patients. The potential mechanisms ­underlying change in each neurometabolite in association with ECT in the context of depression-related functional cir­ cuitry are discussed in the sections that follow.

Glx levels with ECT, which vary by region. Specifically, Glx increased in the sgACC and decreased in the left hippocam­ pus with ECT, and these changes were associated with im­ provements in mood. Earlier cross-sectional 1H-MRS studies have reported deficits of Glx levels within the hippocam­ pus,23 prefrontal cortex36 and ACC,21,22 which point to altered glutamate function in patients with MDD. Observations that N-methyl-d-aspartate (NMDA) receptor antagonists, such as ketamine, can exhibit rapid antidepressant effects also sug­ gest a key role of glutamatergic neurotransmission in rapid antidepressant response.37 The sgACC, which to our know­ ledge has not previously been examined in the context of ECT with 1H-MRS, is implicated in many studies of MDD ­using other neuroimaging modalities. Such investigations re­ port abnormal sgACC function/activity,11 which appear to be modulated with antidepressant treatments, including sleep depravation and deep brain stimulation,38–40 and these find­ ings are compatible with the present observations for ECT. Though not identified as significant in the present study, mean Glx levels in the dACC, like the sgACC, increased across treatment (p = 0.16; Table 3), and this finding is com­ patible with those of our prior report in an independent sam­ ple showing increases in dACC glutamate after 6 ECT ses­ sions as well as results from other earlier studies.18,22 Our observations of joint change in Glx in the sgACC and  left hippocampus with ECT-related clinical response support that variations in Glx contribute, at in least part, to symptom improvements. These findings are in line with the

Controls

C1

n = 33 baseline

Patients

T1

n = 50 baseline

~ 48 hours 2 ECT treatments

~ 4 weeks

T2

n = 42 first follow-up

~ 4 weeks ~ 3 ECT treatments/wk

C2

n = 31 follow-up

T3

n = 33 second follow-up

Glutamate/glutamine The current findings extend previous observations to show disruptions in Glx in patients with MDD and modulation of

10

Fig. 2: Longitudinal and cross-sectional electroconvulsive therapy (ECT) study. C1–C2 = scan times in controls; T1–T3 = scan times in patients.

J Psychiatry Neurosci 2017;42(1)

Treatment and brain chemistry in major depression

neurotransmitter theory of ECT, which posits that changes in monoaminergic as well as glutamatergic signalling act to re­ lieve depressive symptoms.35 At the brain systems level, dif­ ferential effects for Glx in the sgACC and left hippocampus may reflect the frequently reported dissociations between dorsally mediated hyporeactive top–down and ventrally ­mediated hyperreactive bottom–up circuitry in patients with MDD. Specifically, many prior functional imaging studies suggest that hypoactivity in prefrontal cortical networks, to­ gether with elevated activity in ventral limbic structures, con­ tribute to the pathophysiology of MDD.41–44 Here, prefrontal cortical networks are implicated in cognitive control and mood regulation, while ventral networks are considered more closely linked with emotional expression and auto­ nomic responses to emotion; antidepressive treatment has been reported to target these same networks.9,45 The sgACC shares considerable connections with both cortical and sub­ cortical limbic circuitry. It maintains associations with the dACC (part of the dorsal network) via the cingulum, with the hippocampal formation via the precommisural fornix and parahippocmapal gyrus and with the midbrain and brain­ stem (part of the ventral network) via the medial forebrain bundle. Further, the sgACC is adjacent to the septal nuclei, which are reciprocally connected with the hypothalamus, amygdala, olfactory bulb, habenula, cingulate and thala­ mus.46 Altered function of these structures in MDD,7,24 which may include both increased and decreased glutamatergic in­ nervation, may account for variable symptom profiles in MDD and their amelioration by antidepressant treatments,

including ECT. Because of its unique position in the brain, ­Mayberg and colleagues39 first posited the sgACC serves as a conduit between prefrontal mood-regulating regions and subcortical emotional response regions. The decrease in hippocampal Glx toward normative values with ECT, which was different from the pattern observed in the ACC, could potentially reflect a blocking of the excito­ toxic and/or inflammatory effects of long-term exposure to cellular and environmental stressors affecting ventral limbic structures preferentially. Prior evidence supports that en­ hanced glutamate release occurs from exposure to stress and leads to decreased neurogenesis, abnormal dendritic mor­ phology and cell loss in the hippocampus.37 Larger than nor­ mal concentrations of glutamate in ventral limbic structures, specifically in the hippocampus, may thus normalize with ECT to restore glutamatergic homeostasis and neuroprotec­ tive and neurotrophic pathways.

N-acetylaspartate Study results showed decreased NAA in the left hippocam­ pus in patients before treatment relative to controls. At the same time, NAA was also shown to decrease in the dACC and right hippocampus with ECT. NAA is synthesized and stored primarily in neurons. That is, although intercompart­ mental cycling occurs between neurons and glia and changes in NAA have been associated with neurodegenerative white matter disease, because NAA is highly detectable within neur­onal tissue, a change in NAA is generally considered a

Table 2: Demographic and clinical characteristics of study participants Group; mean ± SD (range)* Characteristic Age, yr

MDD, n = 50

Control, n = 33

43.78 ± 14

39.33 ± 12

Sex, no. male:female

23:27

14:19

Adjusted education, yr

15.78 ± 2.70

16.94 ± 2.30

37:12

28:4

Handedness, dextral:nondextral†‡ No. of ECT index sessions§ Lead placement, no. of RUL:mixed:bilateral§

9.96 ± 4.05 (1–22)



34:13:1



Diagnosis, unipolar:bipolar

43:7



No. of responders:remitters¶

19:13



Age at onset, yr† Duration of current episode, yr† Lifetime duration of illness, yr†

27.75 ± 12.23



1.81 ± 2.82



17.78 ± 13.31



Time point

T1, n = 50

T2, n = 42

T3, n = 33

C1, n = 33

C2, n = 31

HAMD-17

24.91 (5.93)**

20.20 (6.23)††

12.46(8.04)‡‡





QIDS-SR

20.53 (4.06)**

16.78(4.87)††

10.27(5.80)‡‡





MADRS

39.04 (9.23)**

32.66 (8.68)††

16.42(11.47)‡‡





C1 = control baseline; C2 = control follow-up; ECT = electroconvulsive therapy; HAMD-17 = Hamilton Rating Scale for Depression; MADRS = Montgomery–Åsberg Depression Rating Scale; QIDS-SR = Quick Inventory of Depressive Symptomatology – Self-Report; RUL = right unilateral lead placement; SD = standard deviation; T1 = patient baseline; T2 = after the second session of ECT; T3 = after the ECT index series. *Unless indicated otherwise. †Data missing for 1 patient. ‡Estimated using the modified Edinburgh Handedness Inventory, where a laterality quotient of < 0.7 defined nondextrals. Data missing for 1 control. §Data missing for 2 patients. ¶Response defined as > 50% improvement in HAMD-17 scores over the course of treatment. Remitters defined as patients with HAMD-17 scores < 7 at the end of ECT. **Significant effect between T1 and T2. ††Significant effect between T2 and T3. ‡‡Significant effect between T1 and T3.



J Psychiatry Neurosci 2017;42(1)

11

Njau et al.

marker of neuronal integrity.47 Reports of hippocampal vol­ ume deficits in patients with MDD48,49 and decreased neuro­ genesis in animal models of depression50 suggest that depres­ sion is linked with neural atrophy. Observations of reduced NAA in patients compared with controls at baseline fits with this hypothesis. However, changes in NAA may also reflect reversible changes in neural metabolism rather than more permanent changes in the number or density of neurons. For example, NAA has been shown to enhance mitochondrial ­energy production from glutamate and to play a role in neur­ onal osmoregulation and axon–glial signalling 51 Observed reductions in NAA may be considered to reflect secondary mechanisms compatible with the anticonvulsant theory of ECT that proposes ECT effects are linked to restor­ ative neural processes that occur in relation to seizure ther­ apy. However, ECT-related changes in NAA may also be compatible with the neurotropic model of ECT. Adult neuro­genesis has been shown to occur after both ECS and epileptic seizures in animals.52,53 Thus, decreases in NAA with ECT may point to changes in the ratio of mature to im­ mature neurons, which may reflect enhanced adult neuro­

Creatine Creatine, which is considered a marker of energy metabo­ lism,47 has been shown to increase in the dACC and sgACC with ECT. The changes that we observed in Cre in the dACC and sgACC may thus suggest a corresponding change in energy use within these brain regions, which may be considered part of the dorsal corticolimbic network. Here, it may be possible that the restorative processes that

B

p = .001 p = .024

5.00

4.50

4.00

3.50

T1 T2 T3 Patients

C

C1 C2 Controls

Error bars: +/- 1 SE

6.10

5.70

p = .001 p = .004

5.30

4.90

4.50

T1 T2 T3 Patients

C1 C2 Controls

D

Error bars: +/- 1 SE

Subgenual anterior cingulate cortex Creatine (AU)

5.50

Dorsal anterior cingulate cortex NAA (AU)

Dorsal anterior cingulate cortex Creatine (AU)

Error bars: +/- 1 SE

p = .001

5.40

p = .025

12.40

p = .002

4.80

p = .050

10.90

4.20

3.60

3.00

Subgenual anterior cingulate cortex Glx (AU)

A

genesis, as has been linked with MDD and treatment re­ sponse in a growing number of studies.50 Notably, studies have also shown that many adult-born neurons within the hippocampal dentate gyrus integrate into pre-existing limbic circuitry to ultimately acquire electrophysiological character­ istics of mature neurons.54,55 Therefore, reductions in NAA may signify prerequisite neural processes that precede and/ or are necessary to stimulate subsequent neurotropic events. Whether ECT-related changes in NAA are linked to changes in cognition may help clarify whether the NAA effects de­ scribed here are beneficial.

T1 T2 T3 Patients

Error bars: +/- 1 SE

C1 C2 Controls

E

9.40

7.90

6.40

T1 T2 T3 Patients

C1 C2 Controls

Error bars: +/- 1 SE

p = .049 p = .016 p = .001 p = .007

4.90

4.40

3.90

3.40

T1 T2 T3 Patients

C1 C2 Controls

5.70

p = .002

Left hippocampus NAA (AU)

8.80

Left hippocampus Glx (AU)

Right hippocampus NAA (AU)

5.40

8.10

7.40

6.70

6.00

T1 T2 T3 Patients

C1 C2 Controls

p = .001

5.30

4.90

4.50

4.10

T1 T2 T3 Patients

C1 C2 Controls

Fig. 3: Cerebrospinal fluid–adjusted 1H-MRS resolved metabolite concentrations in patients scanned at baseline (T1), after the second electroconvulsive therapy (ECT) session (T2) and at the end of the ECT index series (T3) and in controls scanned at 2 time points (C1 and C2). The brackets and p values show significant differences between patients and controls at baseline, significant overall effects of ECT and significant differences between each of 2 time points examined pairwise in patients. Results from follow-up analyses to determine normalization of metabolite concentrations with ECT by comparing T3 and C1 values are not shown. AU = arbitrary units; Glx = glutamate/glutamine; NAA = N-acetylaspartate; SE = standard error.

12

J Psychiatry Neurosci 2017;42(1)

Treatment and brain chemistry in major depression

Table 3: Means and standard errors for cerebrospinal fluid–corrected metabolite concentrations by region, diagnostic group and time point, adjusted for sex and age Patients, time point; mean ± SE Region

Controls, time point; mean ± SE

T1

T2

T3

C1

C2

NAA

5.472 ± 0.111†‡

5.382 ± 0.103†‡

5.086 ± 0.059†‡

5.447 ± 0.091‡

5.466 ± 0.082‡

Glx

9.851 ± 0.174‡

9.867 ± 0.203‡

10.131 ± 0.149‡

9.856± 0.208‡

9.793 ± 0.171‡

Cre

4.645 ± 0.088†‡

4.535 ± 0.090†‡

4.803 ± 0.077†‡

4.473 ± 0.070‡

4.455 ± 0.077‡

Cho

1.164 ± 0.035‡

1.190 ± 0.040‡

1.196 ± 0.035‡

1.183 ± 0.036‡

1.157 ± 0.032‡

Dorsal ACC

Left hippocampus NAA

4.734 ± 0.083*‡

4.905 ± 0.119

4.765 ± 0.089

5.118 ± 0.110*‡

5.107 ± 0.120‡

Glx

8.056 ± 0.174*†‡

7.488 ± 0.193†‡

7.311 ± 0.281†‡

7.377 ± 0.253*‡

7.052 ± 0.305‡

Cre

4.657 ± 0.076‡

4.662 ± 0.077‡

4.717 ± 0.079‡

4.716 ± 0.079‡

4.653 ± 0.069‡

Cho

1.443 ± 0.028‡

1.477 ± 0.036‡

1.440 ± 0.030‡

1.469 ± 0.031‡

1.454 ± 0.029‡

NAA

4.635 ± 0.084†‡

4.680 ± 0.091†‡

4.406 ± 0.063†‡

4.664 ± 0.096‡

4.548 ± 0.107‡

Glx

7.615 ± 0.339‡

7.188 ± 0.260‡

7.563 ± 0.227‡

7.282 ± 0.310‡

6.954 ± 0.277‡

Cre

4.485 ± 0.103‡

4.489 ± 0.101‡

4.838 ± 0.099‡

4.484 ± 0.105‡

4.511 ± 0.136‡

Cho

1.382 ± 0.027‡

1.406 ± 0.025‡

1.368 ± 0.026‡

1.355 ± 0.047‡

1.324 0.032‡

NAA

5.025 ± 0.115‡

4.937 ± 0.113‡

5.144 ± 0.142‡

5.053 ± 0.155‡

5.166 ± 0.075‡

Glx

9.362 ± 0.255*†‡

9.511 ± 0.277‡

10.471 ± 0.384†‡

10.454 ± 0.245*‡

10.266 ± 0.318‡

Cre

4.485 ± 0.103†‡

4.489 ± 0.101†‡

4.838 ± 0.099†‡

4.484 ± 0.105‡

4.511 ± 0.136‡

Cho

1.173 ± 0.032‡

1.136 ± 0.036‡

1.234 ± 0.046‡

1.216 ± 0.043‡

1.266 ± 0.045‡

Right hippocampus

Subgenual ACC

ACC = anterior cingulate cortex; C1 = control baseline; C2 = control follow-up; Cre = creatine; Cho = choline; ECT = electroconvulsive therapy; Glx = glutamate/glutamine; NAA = N-acetylaspartate; SE = standard error; T1 = patient baseline; T2 = after the second session of ECT; T3 = after the ECT index series. *Significant effect between controls and patients at baseline. †Significant main effect of ECT treatment. ‡Corrected for age and sex.

Left hippocampus 7.00

r = 0.29, p = 0.04

ECT-related change in Glx (AU)

ECT-related change in Glx (AU)

5.00

Subegenual anterior cingulate cortex

2.50

.00

–2.50

–5.00 –15.0

.0

15.0

30.0

ECT-related change in HAMD scores

r = 0.44, p = 0.05

4.10

1.20

–1.70

–4.60 –15.0

.0

15.0

30.0

ECT-related change in HAMD scores

Fig. 4: Joint longitudinal changes in Hamilton Rating Scale for Depression (HAMD) scores and glutamate/glutamine (Glx) concentrations in the left hippocampus and subgenual anterior cingulate cortex. Graphs show regression lines and 95% confidence intervals. AU = arbitrary units; ECT = electroconvulsive therapy.



J Psychiatry Neurosci 2017;42(1)

13

Njau et al.

occur after ECT may lead to an increase in the energy re­ quirements in these regions and consequently to increases in Cre. Though variations in Cre appear more likely related to the physiologic rather than the therapeutic effects of ECT (i.e., since associations with change in mood ratings were not observed), it remains possible that such associations are present but occur via intermediary processes.

Limitations H-MRS affords the important opportunity to assay changes in neurochemistry linked with antidepressant response. However, owing to crowded and overlapping resonance frequencies, few 1H-MRS markers are derived from a single compound, rendering it difficult to dissociate some metabo­ lites, such as γ-aminobutyric acid (GABA), without special­ ized pulse sequences or high-field MRI. Consequently, GABA was not measured in this study. Though using a powerful within-subjects research design, attrition in longi­ tudinal assessments served to reduce this advantage such that clinical heterogeneity may still contribute to findings. Here, it is possible that regional changes in 1H-MRS signals may be more sensitive to changes in particular clinical symptoms and/or behavioural outcomes,56 which could be addressed in future studies. The exclusion of some data points because of field inhomogeneities, which occurred more frequently in regions closer to CSF and the cerebral cisterns, is also a limitation. Hence, smaller yet clinically meaningful effects may have remained undetected in the present investigation. For example, since only patients who completed all 3 time points were included for correlational 1

Dorsal anterior cingulate cortex 7.00

Conclusion

r = 0.38, p < 0.05

Baseline NAA (AU)

6.25

5.50

4.75

4.00 –15.0

.0

15.0

30.0

45.0

ECT-related change in HAMD scores

Fig. 5: Baseline N-acetylaspartate (NAA) concentration in the dorsal anter­ior cingulate cortex and longitudinal change in Hamilton Rating Scale for Depression (HAMD) scores. The graph shows the regression line and the 95% confidence intervals. AU = arbitrary units.

14

analyses with mood ratings while all available data were used to examine ECT effects, correlational analyses may be more susceptible to increased type-II error, and more subtle associations between change in metabolite concentration and symptom ratings may have remained undetected. Our findings correlating neurometabolite changes to clinical out­ come were moderate. Further, it is also plausible that such associations are not linear and/or that other mechanisms, such as modulation by GABA or other neurotransmitter systems not measured here and/or different time lapses in the molecular pathways leading to clinical improvement, impact these associations. Though we attempted to mini­ mize potential type-I and type-II error, our findings may yet warrant replication in future studies. Although we ob­ served no interactions, it remains possible that the diagnosis of unipolar versus bipolar depression (14% of the sample) may have impacted our results. Importantly, in line with a new focus on reconceptualizing currently used diagnostic categories by creating a research framework that more accu­ rately represents neurobiological dysfunction in individual patients (i.e., according to National Institute of Mental Health [NIMH] Research Domain Criteria)56 blind to categorical diag­nosis, the primary objective the present study was to ad­ dress links with treatment response in patients experiencing a major depressive episode. However, further studies includ­ ing much larger and more evenly matched diagnostic groups may address potential differential effects between patients with unipolar and bipolar depression specifically. Finally, since the majority of patients received RUL ECT exclusively and bilateral lead placement was based on clinical factors, we were not able to adequately address differential effects of lead placement on the observed results. Future studies may address some of these potential limitations explicitly.

The heterogeneous symptom presentation and clinical course of MDD and the moderate success rate of standard antidepressant therapies underscore the need for biological markers to better inform more personalized treatment deci­ sions. Advancing this goal, the present study demonstrates that ECT elicits pronounced, but differential effects on brain chemistry in cortical and subcortical limbic circuits. Though associations with clinical response are modest, findings sup­ port that the normalization of Glx in the sgACC and hippo­ campus contributes to symptom improvements. Observa­ tions that baseline measures of NAA, a 1H-MRS marker that is typically considered a marker of neural integrity, distin­ guish subsequent response to ECT, suggesting a potential clinical application pending further replication. Cumula­ tively, these findings appear most compatible with a neuro­ transmitter model of ECT response but also align with fea­ tures of other models previously described to explain the mechanisms of ECT. Affiliations: From the Ahmanson-Lovelace Brain Mapping Center, Department of Neurology, University of California, Los Angeles, Calif., USA (Njau, Joshi, Leaver, Vasavada, Woods, Narr); the

J Psychiatry Neurosci 2017;42(1)

Treatment and brain chemistry in major depression

­ epartment of Psychiatry and Biobehavioral Sciences, University of D California, Los Angeles, Calif., USA (Espinoza, Woods, Narr); and the Department of Mathematics, University of Valencia, Valencia, Spain (Marquina). Funding: This study was supported by grant numbers R01MH092301 and K24MH102743 from the National Institute of Mental Health. The content is solely the responsibility of the authors and does not neces­ sarily represent the official views of the National Institute of Mental Health or the National Institutes of Health. Competing interests: None declared.

drome? Meta-analysis of diffusion tensor imaging studies in patients with MDD. J Psychiatry Neurosci 2013;38:49-56. 14. Seminowicz DA, Mayberg H, McIntosh A, et al. Limbic–frontal ­c ircuitry in major depression: a path modeling metanalysis. ­Neuroimage 2004;22:409-18. 15. Fales CL, Barch D, Rundle M, et al. Altered emotional interference processing in affective and cognitive-control brain circuitry in ­major depression. Biol Psychiatry 2008;63:377-84. 16. Kemp AH, Gordon E, Rush A, et al. Improving the prediction of treatment response in depression: integration of clinical, cognitive, psychophysiological, neuroimaging, and genetic measures. CNS Spectr 2008;13:1066-86.

Contributors: R. Espinoza and K. Narr designed the study. S. Njau, A. Leaver, M. Vasavada and K. Narr acquired the data, which S.  Njau, S. Joshi, R. Espinoza, A. Marquina, R. Woods and K. Narr analyzed. S. Njau wrote the article, which all authors reviewed and approved for publication.

17. Joshi SH, Espinoza R, Pirnia T, et al. Structural plasticity of the hippocampus and amygdala induced by electroconvulsive therapy in major depression. Biol Psychiatry 2016;79:282-92.

References

18. Zhang J, Narr K, Woods R, et al. Glutamate normalization with ECT treatment response in major depression. Mol Psychiatry 2013;​ 18:268.

 1. World Health Organization. Depression: a public health concern. 2012. Available: www.who.int/mental_health/management​ /depression​/who_paper_depression_wfmh_2012.pdf.

19. Lyden H, Espinoza R, Pirnia T, et al. Electroconvulsive therapy mediates neuroplasticity of white matter microstructure in major depression. Transl Psychiatry 2014;4:e380.

 2. Krishnan V, Nestler EJ. Linking molecules to mood: new insight into the biology of depression. Am J Psychiatry 2010;167:1305-20.

20. Rao NP, Venkatasubramanian G, Gangadhar B. Proton magnetic resonance spectroscopy in depression. Indian J Psychiatry 2011;53:307.

 3. Phillips ML, Chase H, Sheline Y, et al. Identifying predictors, moderators, and mediators of antidepressant response in major depressive disorder: neuroimaging approaches. Am J Psychiatry 2015;172:124-38.

21. Auer DP, Pütz B, Kraft E, et al. Reduced glutamate in the anterior cingulate cortex in depression: an in vivo proton magnetic resonance spectroscopy study. Biol Psychiatry 2000;47:305-13.

 4. Petrides G, Fink M, Husain M, et al. ECT remission rates in psy­ chotic versus nonpsychotic depressed patients: a report from CORE. J ECT 2001;17:244-53.

22. Pfleiderer B, Michael N, Erfurth A, et al. Effective electroconvul­ sive therapy reverses glutamate/glutamine deficit in the left anter­ ior cingulum of unipolar depressed patients. Psychiatry Res 2003;​ 122:185-92.

 5. Husain MM, Rush A, Fink M, et al. Speed of response and remis­ sion in major depressive disorder with acute electroconvulsive therapy (ECT): a Consortium for Research in ECT (CORE) report. J Clin Psychiatry 2004;65:485-91.

23. Block W, Träber F, von Widdern O, et al. Proton MR spectroscopy of the hippocampus at 3 T in patients with unipolar major depres­ sive disorder: correlates and predictors of treatment response. Int J Neuropsychopharmacol 2009;12:415-22.

 6. Fosse R, Read J. Electroconvulsive treatment: hypotheses about mechanisms of action. Front Psychiatry 2013;4:94.

24. Mu J, Xie P, Yang Z-S, et al. 1 H magnetic resonance spectroscopy study of thalamus in treatment resistant depressive patients. ­Neurosci Lett 2007;425:49-52.

 7. Mayberg HS. Modulating dysfunctional limbic-cortical circuits in depression: towards development of brain-based algorithms for diagnosis and optimised treatment. Br Med Bull 2003;65:193-207.  8. Drevets WC, Price J, Furey M. Brain structural and functional ab­ normalities in mood disorders: implications for neurocircuitry models of depression. Brain Struct Funct 2008;213:93-118.  9. Koenigs M, Grafman J. Prefrontal asymmetry in depression? The long-term effect of unilateral brain lesions. Neurosci Lett 2009;​459:88-90.

25. Vythilingam M, Charles HC, Tupler LA, et al. Focal and lateral­ ized subcortical abnormalities in unipolar major depressive disor­ der: an automated multivoxel proton magnetic resonance spectros­ copy study. Biol Psychiatry 2003;54:744-50. 26. Ende G, Braus D, Walter S, et al. The hippocampus in patients treated with electroconvulsive therapy: a proton magnetic resonance spectro­ scopic imaging study. Arch Gen Psychiatry 2000;57:​937-43.

10. Rogers MA, Kasai K, Koji M, et al. Executive and prefrontal dys­ function in unipolar depression: a review of neuropsychological and imaging evidence. Neurosci Res 2004;50:1-11.

27. Michael N, Erfurth A, Ohrmann P, et al. Neurotrophic effects of electroconvulsive therapy: a proton magnetic resonance study of the left amygdalar region in patients with treatment-resistant de­ pression. Neuropsychopharmacology 2003;28:720-5.

11. Drevets WC, Savitz J, Trimble M. The subgenual anterior cingulate cortex in mood disorders. CNS Spectr 2008;13:663.

28. Hamilton M. A rating scale for depression. J Neurol Neurosurg ­Psychiatry 1960;23:56-62.

12. Lorenzetti V, Allen N, Fornito A, et al. Structural brain abnormal­ ities in major depressive disorder: a selective review of recent MRI studies. J Affect Disord 2009;117:1-17.

29. Montgomery SA, Asberg M. A new depression scale designed to be sensitive to change. Br J Psychiatry 1979;134:382-9.

13. Liao Y, Huang X, Wu Q, et al. Is depression a disconnection syn­

30. Hess AT, Tisdall D, Andronesi O, et al. Real‐time motion and B0



J Psychiatry Neurosci 2017;42(1)

15

Njau et al.

corrected single voxel spectroscopy using volumetric navigators. Magn Reson Med 2011;66:314-23. 31. Tisdall MD, Hess A, Reuter M, et al. Volumetric navigators for prospective motion correction and selective reacquisition in neuro­ anatomical MRI. Magn Reson Med 2012;68:389-99. 32. Joshi S, Njau S, Marquina A, et al. Denoising of MR spectroscopy signals using total variation and iterative Gauss-Siedel gradient updates. IEEE International Symposium on Biomedical Imaging; 2015 Apr 16 -19; New York, New York. 33. Shattuck DW, Leahy R. BrainSuite: an automated cortical surface identification tool. Med Image Anal 2002;6:129-42. 34. Gussew A, Erdtel M, Rzanny R, et al. Absolute quantitation of brain metabolites with respect to the CSF fraction in 1H-MR spec­ troscopic volumes. MAGMA 2012;25:321-33. 35. Kellner CH, Greenberg R, Murrough J, et al. ECT in treatment-­ resistant depression. Am J Psychiatry 2012;169:1238-44. 36. Hasler G, van der Veen J, Tumonis T, et al. Reduced prefrontal glutamate/glutamine and γ-aminobutyric acid levels in major de­ pression determined using proton magnetic resonance spectros­ copy. Arch Gen Psychiatry 2007;64:193-200. 37. Sanacora G, Treccani G, Popoli M. Towards a glutamate hypothesis of depression: an emerging frontier of neuropsychopharmacology for mood disorders. Neuropharmacology 2012;62:63-77.

treatment response in depression: integration of clinical, cognitive, psychophysiological, neuroimaging, and genetic measures. CNS Spectr 2008;13:1066-86. 44. Price JL, Drevets WC. Neurocircuitry of mood disorders. Neuro­ psychopharmacology 2010;35:192-216. 45. Clark CP, Brown G, Archibald S, et al. Does amygdalar perfusion correlate with antidepressant response to partial sleep deprivation in major depression? Psychiatry Res 2006;146:43-51. 46. Nieuwenhuys R, Voogd J, Van Huijzen C. The human central nervous system: a synopsis and atlas. 4th ed. Berlin: New York Springer; 2007. 47. Yildiz-Yesiloglu A, Ankerst D. Review of 1 H magnetic resonance spectroscopy findings in major depressive disorder: a meta-­ analysis. Psychiatry Res 2006;147:1-25. 48. Sheline YI, Mittler B, Mintun M. The hippocampus and depres­ sion. Eur Psychiatry 2002;(Suppl 3):300-5. 49. Joshi SH, Espinoza R, Pirnia T, et al. Structural plasticity of the hippocampus and amygdala induced by electroconvulsive therapy in major depression. Biol Psychiatry. In press. 50. Eisch AJ, Petrik D. Depression and hippocampal neurogenesis: a road to remission? Science 2012;338:72-5. 51. Baslow MH. N-acetylaspartate in the vertebrate brain: metabolism and function. Neurochem Res 2003;28:941-53.

38. Gillin JC, Buchsbaum M, Wu J, et al. Sleep deprivation as a model experimental antidepressant treatment: findings from functional brain imaging. Depress Anxiety 2001;14:37-49.

52. Chen F, Madsen T, Wegener G, et al. Repeated electroconvulsive seizures increase the total number of synapses in adult male rat hippocampus. Eur Neuropsychopharmacol 2009;19:329-38.

39. Mayberg HS, Lozano A, Voon V, et al. Deep brain stimulation for treatment-resistant depression. Neuron 2005;45:651-60.

53. Schmidt HD, Duman R. The role of neurotrophic factors in adult hippocampal neurogenesis, antidepressant treatments and ani­ mal models of depressive-like behavior. Behav Pharmacol 2007;18:​ 391-418.

40. Nobler MS, Oquendo M, Kegeles L, et al. Decreased regional brain metabolism after ECT. Am J Psychiatry 2001;158:305-8. 41. Mayberg HS, Liotti M, Brannan S, et al. Reciprocal limbic-cortical function and negative mood: converging PET findings in depres­ sion and normal sadness. Am J Psychiatry 1999;156:675-82. 42. Drevets WC, Price JL, Furey ML. Brain structural and functional abnormalities in mood disorders: implications for neurocircuitry models of depression. Brain Struct Funct 2008;213:93-118. 43. Kemp AH, Gordon E, Rush A, et al. Improving the prediction of

16

54. Lledo P-M, Alonso M, Grubb M. Adult neurogenesis and func­ tional plasticity in neuronal circuits. Nat Rev Neurosci 2006;7:​ 179-93. 55. Cuthbert BN. The RDoC framework: facilitating transition from ICD/DSM to dimensional approaches that integrate neuroscience and psychopathology. World Psychiatry 2014;13:28-35. 56. Oldfield RC. The assessment and analysis of handedness: the Ed­ inburgh Inventory. Neuropsychologia 1971;9:97-113.

J Psychiatry Neurosci 2017;42(1)

Neurochemical correlates of rapid treatment response to electroconvulsive therapy in patients with major depression.

Electroconvulsive therapy (ECT) is a highly effective brain stimulation treatment for severe depression. Identifying neurochemical changes linked with...
1MB Sizes 0 Downloads 9 Views